PREPARE_TRANSACTION(7)

💖 PREPARE TRANSACTION 💖

okay so bestie, we're in the PostgreSQL 18.6 Documentation manual, section 7, and today's episode is PREPARE_TRANSACTION — the glow-up that gets your current transaction ready for two-phase commit. yes, the actual commit comes later. we're just doing the ✨prep✨.

---

🌸 NAME 🌸

PREPARE_TRANSACTION — prepare the current transaction for two-phase commit. it's giving "getting ready backstage" energy fr fr.

---

🎀 SYNOPSIS 🎀

PREPARE TRANSACTION transaction_id

---

🦋 DESCRIPTION 🦋

so PREPARE TRANSACTION prepares the current transaction for two-phase commit. that's the whole vibe.

after you run it, the transaction is no longer associated with the current session. instead, its state is fully stored on disk — like, fully fully — and there is a very high probability that it can be committed successfully, even if a database crash happens before the commit is ever requested. she's durable, she's safe, she's prepared. 💅

once a transaction is prepared, you can later commit or roll it back with COMMIT PREPARED or ROLLBACK PREPARED, respectively. and here's the cute part: those commands can be issued from any session, not only the one that executed the original transaction. so your bestie can finish what you started, no cap.

from the point of view of the issuing session, PREPARE TRANSACTION is not unlike a ROLLBACK command. after executing it, there is no active current transaction, and the effects of the prepared transaction are no longer visible. 😭 (the effects will become visible <em>again</em> if the transaction is committed. so like, it's not gone gone. it's just backstage.)

and if the PREPARE TRANSACTION command fails for any reason, it becomes a ROLLBACK: the current transaction is canceled. rip. 💀

---

✨ PARAMETERS ✨

• transaction_id — an arbitrary identifier that later identifies this transaction for COMMIT PREPARED or ROLLBACK PREPARED. it's basically the transaction's @ handle. the rules tho, bestie:
  • the identifier must be written as a string literal
  • it must be less than 200 bytes long
  • it must not be the same as the identifier used for any currently prepared transaction (no duplicate handles, that's not the vibe)

---

👑 NOTES 👑

okay real talk. PREPARE TRANSACTION is not intended for use in applications or interactive sessions. its whole purpose is to allow an external transaction manager to perform atomic global transactions across multiple databases or other transactional resources. so unless you're literally writing a transaction manager, you probably shouldn't be using PREPARE TRANSACTION. i said what i said. 🙈

this command must be used inside a transaction block. use BEGIN to start one. no block, no prep, periodt.

also, it is not currently allowed to PREPARE a transaction that has:

• executed any operations involving temporary tables or the session's temporary namespace
• created any cursors WITH HOLD
• executed LISTEN, UNLISTEN, or NOTIFY

those features are too tightly tied to the current session to be useful in a transaction to be prepared. they're codependent, it's not healthy, we don't do that here. 🚫

if the transaction modified any run-time parameters with SET (without the LOCAL option), those effects persist after PREPARE TRANSACTION, and will not be affected by any later COMMIT PREPARED or ROLLBACK PREPARED. so in this one respect, PREPARE TRANSACTION acts more like COMMIT than ROLLBACK. plot twist.

all currently available prepared transactions are listed in the pg_prepared_xacts system view. that's your little tracker, bestie. check it. 👀

⚠️ Caution ⚠️

it is unwise to leave transactions in the prepared state for a long time. this will interfere with the ability of VACUUM to reclaim storage, and in extreme cases could cause the database to shut down to prevent transaction ID wraparound (see Section 24.1.5). 💀💀💀

keep in mind also that the transaction continues to hold whatever locks it held. the intended usage of the feature is that a prepared transaction will normally be committed or rolled back as soon as an external transaction manager has verified that other databases are also prepared to commit. that's the whole point. don't leave her on read.

if you have not set up an external transaction manager to track prepared transactions and ensure they get closed out promptly, it is best to keep the prepared-transaction feature disabled by setting max_prepared_transactions to zero. this will prevent accidental creation of prepared transactions that might then be forgotten and eventually cause problems. set it to zero. do it for me. 🫶

---

🌈 EXAMPLES 🌈

prepare the current transaction for two-phase commit, using foobar as the transaction identifier:

PREPARE TRANSACTION 'foobar';

look at her go. iconic. ✨

---

💜 COMPATIBILITY 💜

PREPARE TRANSACTION is a PostgreSQL extension. it is intended for use by external transaction management systems, some of which are covered by standards (such as X/Open XA), but the SQL side of those systems is not standardized. so like, the vibe is standardized, the syntax is not. iykyk.

---

🌟 SEE ALSO 🌟

• COMMIT PREPARED (COMMIT_PREPARED(7))
• ROLLBACK PREPARED (ROLLBACK_PREPARED(7))

---

and that's the tea, bestie 🍓 go prepare responsibly — this was PostgreSQL 18.6, dated 2026, straight from the PREPARE TRANSACTION(7) page. mwah. 💖🎀🦋🌸💅